Sheralee Brindell

Within that portion of philosophy of science concerned with the nature of scientific explanation, has emerged a fundamental disagreement about the nature of explanatory relevance as a primary source of explanatory power. And while everyone agrees that being relevant within an explanation requires that the explaining fact have something to do with the fact to be explained, there is little agreement about what this means. As the debate about the nature of explanation has developed within different metaphysical frameworks, various answers to the "what is explanatory relevance?" question have been offered. Hempel and Oppenheim's D-N and I-S accounts, for example, construed relevance as a function of being an instance of a law appearing as part of an explanans. Wesley Salmon's objections to this account of scientific explanation would require that relevance be located within the objective causal nexus underlying the events to be explained. Still other theorists analyze relevance epistemically, in terms of an answer's responsiveness to a question asked. But as Salmon and others rightly point out, such analyses are fraught with difficulty--none of these epistemic analyses appear able to provide adequate criteria for an answer's correctness . In order to get a clear view of the problems of relevance within an epistemic/erotetic account, I present two of the leading erotetic models--one semantic, the other speech-act theoretic. I then distill the most pointed realist objections against them and show how both suffer from the same structural difficulty. As a way toward acquiring a fresh perspective, I then present a third erotetic account and argue that its conceptions of 'problem-situations' and explanatory requests suggest a line of analysis previously unexplored. I proceed to show that an inquirer's status as epistemically dependent, her question as a request for help, and her need to trust her interlocutor, all suggest that recasting the problems of relevance into moral terms may resolve these difficulties as they are epistemically conceived. Finally, I argue that a respondent's trustworthiness is capable of underwriting the necessary epistemic claims, and proceed to develop a set of criteria according to which one may qualify as trustworthy to explain. I conclude by showing how a respondent's trustworthiness addresses the difficulties present within each of the three erotetic accounts examined, and argue that this result suggests that trustworthiness is fundamental to an erotetic conception of explanation

This chapter argues that doubts about the scientific status of the field sciences often rest on mistaken preconceptions about the nature of the evaluative relation between empirical evidence and hypothesis or theory, namely, that it is some sort of formal logical relation. It argues that there is a potentially more fruitful approach to understanding the nature of the support offered by empirical evidence to scientific hypotheses. The first part of the chapter briefly reviews the traditional philosophical take on the scientific method in order to clarify its most serious problems. It shows that these problems are greatly exacerbated when science moves from the artificially controlled environment of the laboratory to the messy uncontrollable world of nature. The second part ferrets out some highly general, causal components in the methodological reasoning of nonhistorical field scientists. It argues that differences in patterns of evidential reasoning in the experimental sciences versus the field sciences, and in the historical versus nonhistorical field sciences, seem tailored to pervasive causal differences in their epistemic situations.